SEED Research & Announcements Blogs Publications Open Source Careers Contact Us Research & Announcements Blogs Publications Open Source Careers Contact Us

Project PICA PICA

SEED

PICA PICA is a real-time raytracing experiment featuring self-learning agents in a procedurally-assembled world. Built using SEED's R&D engine Halcyon, this AI playground showcases real-time raytracing using DirectX Raytracing, Microsoft's newly announced API which enables developers to bring raytracing to life.

This demo showcases several novel approaches developed specifically for real-time raytracing, including raytraced reflections, ambient occlusion, soft shadows, transparency, translucency and subsurface scattering, as well as dynamic global illumination.

Here are a few screenshots - click to see them at full resolution!

PICA PICA Screenshot 01 PICA PICA Screenshot 02PICA PICA Screenshot 03 PICA PICA Screenshot 04

Make sure to attend this year's GDC, where several techniques from PICA PICA will be presented in the following talks.
You can also see it running live at the conference's expo, at NVIDIA's booth.

 

Related News

Improving Generalization in Game Agents with Imitation Learning

SEED
Jul 16, 2024
How do we efficiently train in-game AI agents to handle new situations that they haven’t been trained on?

Towards Optimal Training Distribution for Photo-to-Face Models

SEED
Jul 8, 2024
How do we best construct game avatars from photos? This presentation discusses a work in progress with an optimized view of the training data.

Incorporating ML Research Into Audio Production: ExFlowSions Case Study

SEED
Jun 25, 2024
Mónica Villanueva and Jorge García present the challenges and lessons learned from turning a machine learning generative model from a research project into a game production tool.